June 9, 2009 Letters to the Editor

Updated: June 14, 2009.

Effect of Taxes on HSWR

I received this letter from Michael.

Hi John - A question I've been thinking about and meaning to ask about for awhile - [what] do you think about the effect of taxes on HSWR [Historical Surviving Withdrawal Rates]? It was brought home to me again with your analysis of the Balanced Fund vs. S&P/TIPS. While the former had higher HSWR at virtually all levels of 100E/P, how much of the difference would be taken up by the presumably higher taxes on it? I imagine the precise calculation is virtually impossible as you would have to know the sources of income in the Balanced fund each year (short term capital gain/loss, long term, ordinary income, and dividend) and that year's tax rate for each source, but IN GENERAL I would imagine the income from the Balanced Fund would be more ordinary in nature than the S&P/Tips (e.g., higher taxes) AND the S&P/TIPS would have more long term gains than the Balanced Fund (e.g., lower taxes). Thoughts?

More generally, as your site is call "Early" Retirement, I would be curious how you would think about the effect of taxes assuming "Early" retirement vs. more traditional retirement ages in the context of what the ages of allowable withdrawals from 401(k) / IRA plans? E.g., assuming "Early" retirement means more reliance on non-401(k), non-IRA savings accounts, at least earlier in retirement.

Thanks.

HERE IS MY RESPONSE

Thank you once again for an excellent letter.

You hit upon the reason that I do not include the effect of taxes. It is hard work. It is especially hard if you want it to be meaningful to individuals.

To a certain extent, I can hide behind the fact that internally generated income in retirement accounts is not taxed before withdrawals. The adjustment is in determining the amount of retirement income needed each year.

But the real reason is that it takes a lot of effort to produce anything meaningful.

I do not know which portfolio would be more efficient after taxes. Remember that the inflation adjustments of TIPS are taxed immediately.

Accumulators face a very difficult time because the tax authorities hold them back, reducing returns in taxable accounts. Many early retirees, however, save large amounts over a short number of years. The investment gains do not have that much of an effect. In fact, avoiding losses during accumulation is an important issue those who retire very early.

Those who are able to reach large balances in their retirement accounts only have to have enough in taxable accounts to make it to age 59.5 (IRAs) and possibly earlier. This eases the problem somewhat. Working part time for five or ten years can permit you to gain most of the advantages of retirement even earlier, especially if it comes with health care benefits.

Two additional things

Michael followed up almost immediately with these additional thoughts.

Hi John - as a follow on to my question regarding taxes on the Balanced Fund vs. S&P/TIPS HSWR analysis:

Another well-known "Balanced Fund" with a long track record that would be an interesting to compare is VWELX (Vanguard Wellington) - around since 1929. Whereas FKINX is biased toward fixed income (~60 FI, ~40 equities), VWELX tends to be biased toward equities (~60, with ~40 FI). But importantly, it is Balanced and has a long track record to compare. I SUSPECT the HSWR graph as a function of 100E/P would fall somewhere between FKINX and S&P/TIPS but that is just a guess (and for the lawyers among us, I have NO connection whatsoever to VWELX).

Also, when we talk about HSWR [Historical Surviving Withdrawal Rates], we are talking about the future and making probabilistic analysis based upon valuations which determine safe, reasonably safe, 50/50, likely failure, etc. However with the analysis on FKINX, we are looking at the PAST and know exactly what the path of values was. So, is the SWR, then, the withdrawal amount which leaves you with exactly a zero balance at then end of 30 yrs? [What about an] amount that leaves you with your original balance, adjusted for inflation?

Thanks again.

HERE IS MY RESPONSE

Thank you again.

Thank you for identifying some worthwhile additions to Current Research O.

I suspect that VWELX would do very well in any comparison. It might easily be the best choice. I do not know. Right now, I do not have the relevant data (annual percentages of nominal total returns).

The Historical Surviving Withdrawal Rate has a balance of zero at the end of the 30 years. I have used the term Constant Terminal Value Rate for a final balance equal to the initial balance (plus inflation). I will go ahead and calculate them for FKINX and the S&P500/TIPS combination.

I was able to prove a simple, but useful theorem. The withdrawal rates at intermediate final balances have the same proportions as between the final balances. That is, for a final balance of 50% of the initial balance (plus inflation), simply split the difference between the zero balance withdrawal rate and the constant terminal value withdrawal rate.

You can make an excellent estimate of the Safe Withdrawal Rate by drawing a line parallel with that in the graph (which is the calculated rate, the most likely result). Set it so that one or two data points are outside the line and the rest are all inside.

This graphical procedure is amazing accurate. [I made many comparisons with standard deviation calculations in my early work.] It handles unsymmetrical data as well. It does not make assumptions about the underlying distribution. It is reasonably accurate when compared to the standard approximation (Gaussian distribution for the inner portion of the data).

Repeat this graphical procedure on the high side to determine the Unsafe Rate. Split the differences and you have good approximations of the 20% and 80% probabilities.

Accumulating dividend producing stocks

I received this letter from David Shafer. Dave’s web site is:

David Shafer’s Web Site

Love your site. I have a little less than 20 years to go before retirement. Recently, I have come to believe that along with investing for capital growth, I should invest for income. As a value investor, a Buffett disciple, and someone who can find their way around financial reports I don't worry about diversification. I have found several offerings [a REIT and a Master Limited Partnership] that have a history of good dividend growth as well as good current dividend yields indicating attractive valuations. They are paying their dividends out of free cash flow and have good coverage ratio's. It occurs to me that investing for income is far easier to track if your goal is retirement income than worrying about absolute value and safe withdrawal rates. I have created a spreadsheet that tracks income, my yield, and a couple of financial metrics for each company. It is really easy to see how much income is being produced [even though I am currently re-investing the dividends]. You can do the same analysis with real estate investing or even TIPS, Bank CDs or bonds. Doesn't it make more sense to invest this way with at least a portion of your portfolio? You know what your income goal is, you simply track how close to it you are every couple of years and adjust your investing accordingly.

It just seems like you are comparing apples with oranges when trying to look at portfolio value and extrapolate to income! When my wife asks me how we are doing I can give her a completely accurate assessment! We have a net worth of XXXX currently producing XXXXX of annual income....As I get closer to retirement I can move some of my Berkshire Hathaway over to dividend producing instruments based on the relative price of Berkshire, effectively selling high. I don't have to worry about stock market swoons right before or during retirement because I have those dividends coming in.Just a thought........

HERE IS MY RESPONSE

Thank you. You are right.

I have opened up a new section Dividend Guidelines [button on left] in response to this letter.

You are doing exactly the right thing.

Real estate is especially attractive at this time. Prices have fallen and mortgage rates are low. Inflation is likely roar within the next couple of years. The only reason we don’t see high inflation in the official numbers now is because housing prices (and, therefore, equivalent rents) have been falling. This won’t last long. People flock to tangibles such as real estate and commodities at times of high inflation.

Keep in mind that you will still have some high dollar purchases to make in retirement. You need to pay attention to asset prices as well as income. But income is far more important.

Dividend Guidelines

June 9, 2009 Letters to the Editor

Effect of Taxes on HSWR

Two additional things

Accumulating dividend producing stocks

Letters to the Editor in 2009

Letters to the Editor in 2008

Letters to the Editor in 2007

Letters to the Editor in 2006

Letters to the Editor in 2005